Unlimited OCR vs PaddleOCR (PP-OCR)
Pick Unlimited OCR to turn long, multi-page documents such as reports and books into Markdown with block types and boxes in one request. Pick PaddleOCR (PP-OCR) to read text lines from receipts, signs and screenshots on CPU, mobile or in the browser. The main split is scope: Unlimited OCR parses whole documents on a GPU, while PP-OCR is a small detector and recognizer that returns lines and boxes.
Unlimited OCRbaidu/unlimited-ocr | PaddleOCR (PP-OCR)paddle/paddleocr | |
|---|---|---|
| Type | Open source | Open source |
| Price | 2 credits / page | 1 credit / page |
| Per 1,000 pages | $2.00 | $1.00 |
| Typical speed | 1.2 s / page | 0.7 s / page |
| Best for | Books and very long documents | Fast printed text, Chinese and English |
| Inputs | PDF, images | PDF, images |
| Languages | 30+ languages | 80+ languages |
| Max pages per request | 1,000 | 500 |
| Tables | ✓ Yes | No |
| Handwriting | No | No |
| Math | ✓ Yes | No |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | No | ✓ Yes |
| Forms | No | No |
| Multilingual | ✓ Yes | ✓ Yes |
| On OCRRouter | Available | Available |
Prices are what you pay on OCRRouter. Speed is the provider's typical figure, not a benchmark we ran.
Which to choose
Choose Unlimited OCR if
- You need full document to Markdown. It parses 40+ pages in a single forward pass and emits block types and bounding boxes alongside text.
- You are building bulk PDF to text for RAG over long reports and books, with memory kept roughly constant across pages.
- You have an NVIDIA GPU with CUDA 12.9 or 13.0 and want an MIT licence with a hosted option in Baidu Cloud's OCR product.
Choose PaddleOCR (PP-OCR) if
- You have no GPU. PP-OCR runs on CPU, NVIDIA GPU, Kunlunxin XPU and other accelerators, with a reported 5.2x CPU speedup using OpenVINO.
- You do edge, mobile or in-browser OCR. The PP-OCRv6 medium model is 34.5M parameters, and the PaddleOCR.js SDK runs in the browser.
- You read printed Chinese, English or Japanese text, or any of 46 Latin-script languages, with one PP-OCRv6 model covering 50 languages.
Key differences
| Unlimited OCR | PaddleOCR (PP-OCR) | |
|---|---|---|
| Scope | Multi-page document parser; Markdown, plain text and bounding boxes | Text detection and recognition; plain text, bounding boxes and JSON |
| Model size | 3B total, about 500M active per token (Unlimited OCR 2 is 3.34B) | PP-OCRv6 medium: 34.5M |
| Hardware | NVIDIA GPU with CUDA 12.9 or 13.0; version 2 under SGLang runs out of memory on 12 GB GPUs; not for CPU-only machines | CPU, NVIDIA GPU, Kunlunxin XPU and other accelerators; 0.13s on A100 reported |
| Licence | MIT | Apache-2.0 |
| How you run it | Self-host with vLLM, SGLang or transformers, or use the hosted endpoint in Baidu Cloud | Local paddleocr Python package, C++ or the PaddleOCR.js browser SDK |
| Tables and formulas | Not published | Not handled by PP-OCR; needs the separate PP-Structure or VL pipelines |
| Known failure | Text repetition and looping on some pages, hallucinated text, lost bold and italic styling; context capped at 32,768 tokens | PP-OCRv6 weights contain denormals that make CPU inference 4 to 11x slower in RapidOCR |
Sources
- Hugging Face model card
- GitHub repo
- Issue #55 repetition
- PaddleOCR GitHub README
- PP-OCRv6 on Hugging Face
- RapidOCR issue #743 denormal weights
What each one is
Unlimited OCR
Open-source model from Baidu that reads hundreds of pages in one pass with flat memory use.
Unlimited OCR detailsPaddleOCR (PP-OCR)
Baidu's classic open-source text detection and recognition pipeline. Light enough to run on CPU.
PaddleOCR (PP-OCR) detailsTry both on your own file
Run the file once with each model and compare the output.
Use one, fall back to the other
{"model": "baidu/unlimited-ocr","fallbacks": ["paddle/paddleocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Unlimited OCR alternativesPaddleOCR (PP-OCR) alternativesBest OCR models in 2026